Open msa7 opened 6 years ago
thanks for the feedback. I will investigate this.
Just to verify, does running your containers without sentry drastically reduce the hyperkit cpu?
does running your containers without sentry drastically reduce the hyperkit cpu?
yes
I found a workaround. Run sentry on Mac. On file change rerun docker container, like this:
bin/sentry \
-w "./src/**/*.cr" -w "./src/**/*.ecr" \
-b "" -r "docker-compose" --run-args "run --rm test crystal spec --debug"
Please also consider https://github.com/samueleaton/sentry/pull/25
so sentry is not a cpu hog when being run directly on the host? its only when running in docker?
What is the processor and ram on your Mac? I am running a few sentries locally and hyperkit is staying below 20% (usually bellow 10%).
I'm having the same issue, getting 300%+ (4 cores) with sentry running in the container. (lucky framework uses sentry for watch
).
Related: https://github.com/docker/for-mac/issues/1094
Though for me a factory reset doesn't solve the problem.
To answer your question to msa7:
MacBook Pro (Retina, 15-inch, Mid 2014) 2.5 GHz Intel Core i7 16 GB 1600 MHz DDR3
Docker has access to: 4 CPUs 6 GiB of Memory 1 GiB of Swap
I suppose I could increase the Swap? For now planning on running watch from host and triggering a recompile in the container from there like msa7.
OK a few notes:
modified
timestamps) is going to be bad in docker for mac until they fix it.@KCErb yes please let me know what you find.
Thanks for the useful shard.
I run three Sentry, each in the docker container.
I have huge CPU load
Maybe Senty should use
kqueue
or similar, like rerun gem?